Composition of Experts: A Modular Compound AI System Leveraging Large Language Models

TL;DR

This paper presents the Composition of Experts (CoE), a modular AI system using multiple expert LLMs with dynamic routing, improving performance and efficiency over monolithic models.

Contribution

It introduces a novel two-step routing approach for training and deploying modular LLM systems, addressing scalability and resource utilization challenges.

Findings

Achieved superior performance with reduced computational overhead.

Demonstrated effectiveness on benchmark tasks with open weight LLMs.

Implemented efficient system architecture leveraging SambaNova hardware.

Abstract

Large Language Models (LLMs) have achieved remarkable advancements, but their monolithic nature presents challenges in terms of scalability, cost, and customization. This paper introduces the Composition of Experts (CoE), a modular compound AI system leveraging multiple expert LLMs. CoE leverages a router to dynamically select the most appropriate expert for a given input, enabling efficient utilization of resources and improved performance. We formulate the general problem of training a CoE and discuss inherent complexities associated with it. We propose a two-step routing approach to address these complexities that first uses a router to classify the input into distinct categories followed by a category-to-expert mapping to obtain desired experts. CoE offers a flexible and cost-effective solution to build compound AI systems. Our empirical evaluation demonstrates the effectiveness of CoE in achieving superior performance with reduced computational overhead. Given that CoE comprises of many expert LLMs it has unique system requirements for cost-effective serving. We present an efficient implementation of CoE leveraging SambaNova SN40L RDUs unique three-tiered memory architecture. CoEs obtained using open weight LLMs Qwen/Qwen2-7B-Instruct, google/gemma-2-9b-it, google/gemma-2-27b-it, meta-llama/Llama-3.1-70B-Instruct and Qwen/Qwen2-72B-Instruct achieve a score of $59.4$ with merely $31$ billion average active parameters on Arena-Hard and a score of $9.06$ with $54$ billion average active parameters on MT-Bench.